kola-openstack fails due to exceeding quota

[marmijo]

The kola openstack job has been failing lately on every stream due to exceeding quota. We get the following errrors:

[2023-07-11T17:57:50.354Z]         harness.go:1704: Cluster failed starting machines: waiting for instance to run: Server 
reported ERROR status: {500 2023-07-11 17:57:40 +0000 UTC  Build of instance 97deb130-396d-438d-87e3-706114227adc
 aborted: VolumeSizeExceedsAvailableQuota: Requested volume or snapshot exceeds allowed gigabytes quota. Requested
 10G, quota is 1600G and 1600G has been consumed.}

[2023-07-11T17:57:52.851Z] 2023-07-11T17:57:52Z kola: retryloop: failed to bring up machines: waiting for instance to run: 
Server reported ERROR status: {500 2023-07-11 17:57:47 +0000 UTC  Build of instance 897854a4-7a4d-4b72-953c-
8b95b0c387de aborted: VolumeSizeExceedsAvailableQuota: Requested volume or snapshot exceeds allowed gigabytes quota. 
Requested 10G, quota is 1600G and 1600G has been consumed.}

or

[2023-07-14T19:35:14.192Z] 2023-07-14T19:35:13Z kola: Flight failed: Request forbidden: [POST
https://compute.public.mtl1.vexxhost.net/v2.1/360b86ddb1994b55a7c5757f8adc3637/os-keypairs], error message: 
{"forbidden": {"code": 403, "message": "Quota exceeded, too many key pairs."}}

Volumes are being created and are not being cleaned up causing the quota exceeded errors. We have been able to delete them manually and get a few tests to pass, but the volumes don't get cleaned up so eventually tests begin to fail. As as result, the entire kola-openstack job fails.

[marmijo]

Let's snooze kola-openstack to unblock the FCOS pipeline while we investigate this. I put up a PR to snooze the cloud test until the end of July. #890

[dustymabe]

for the too many keypairs thing I figured out what the problem was. The web interface was telling me there were no keypairs but the CLI told me a different story (there were a lot). I cleaned them up and created this issue: coreos/coreos-assembler#3550